3 ms·

I think you're accidentally making a jump in meta-levels: The equivalent of an existing handwriting recognizer for e.g. RSA wouldn't be a model that tells you

I think you're accidentally making a jump in meta-levels: The equivalent of an existing handwriting recognizer for e.g. RSA wouldn't be a model that tells you how to crack RSA, it'd be a model that _does_ crack it (maybe by returning a probability distribution over the plaintext that's better than uniform). That feels pretty unlikely to me personally, but maybe that's doable, who knows. For the standard of "the LLM tells us how to crack RSA," it feels like the equivalent for handwriting recognition would be the LLM outputting a description of a novel algorithm for handwriting recognition (and probably one that isn't just an existing one with some hyperparameters tweaked). My understanding of the weaknesses of current LLMs is that they're actually pretty bad at that, since they tend to regurgitating existing content when they're able to.

bogwog 3y ago
> The equivalent of an existing handwriting recognizer for e.g. RSA wouldn't be a model that tells you how to crack RSA, it'd be a model that _does_ crack it (maybe by returning a probability distribution over the plaintext that's better than uniform). That feels pretty unlikely to me personally, but maybe that's doable, who knows. Well, my point was based on the assumption that there is a flaw/backdoor in RSA, but nobody has discovered it yet. So you'd just need a model that can take a textual description of the algorithm as input, and spit out a list of logical errors/flaws. I'm not saying there is a flaw or backdoor, but if there were, then a LLM would potentially be able to find it, while a team of human experts could miss it. A model that receives an RSA encrypted payload and outputs the decrypted version would of course be impossible unless the above assumption is true... or you give it access to some compute power so it can either try to brute-force it, or try to track down and hack the servers with the keys :P
- remexre 3y ago
  > So you'd just need a model that can take a textual description of the algorithm as input, and spit out a list of logical errors/flaws. Right, I'm saying that this is the thing that's dissimilar from handwriting recognition -- the nearer equivalent would an LLM creating a textual description of a novel approach to handwriting recognition, which as far as I'm aware, is beyond current models.

bogwog 3y ago

> The equivalent of an existing handwriting recognizer for e.g. RSA wouldn't be a model that tells you how to crack RSA, it'd be a model that _does_ crack it (maybe by returning a probability distribution over the plaintext that's better than uniform). That feels pretty unlikely to me personally, but maybe that's doable, who knows. Well, my point was based on the assumption that there is a flaw/backdoor in RSA, but nobody has discovered it yet. So you'd just need a model that can take a textual description of the algorithm as input, and spit out a list of logical errors/flaws. I'm not saying there is a flaw or backdoor, but if there were, then a LLM would potentially be able to find it, while a team of human experts could miss it. A model that receives an RSA encrypted payload and outputs the decrypted version would of course be impossible unless the above assumption is true... or you give it access to some compute power so it can either try to brute-force it, or try to track down and hack the servers with the keys :P

remexre 3y ago
> So you'd just need a model that can take a textual description of the algorithm as input, and spit out a list of logical errors/flaws. Right, I'm saying that this is the thing that's dissimilar from handwriting recognition -- the nearer equivalent would an LLM creating a textual description of a novel approach to handwriting recognition, which as far as I'm aware, is beyond current models.